Context Graph Agent

Local code graphs for AI coding agents

6 followers

Local code graphs for AI coding agents

6 followers

CGA gives AI coding agents focused repository evidence instead of broad prompt dumps. It indexes files, symbols, calls, imports, dependencies, and lightweight data flow into a local graph, then exposes MCP-compatible tools for retrieval before generation. The admin UI, 3D graph viewer, CGA-Relay, schedules, backup, and benchmark workflow help teams test local-first context quality.

Free

Launch tags:Open Source•Developer Tools•Artificial Intelligence

Launch Team

ClarifyThe autonomous CRM that helps you sell more

Promoted

Maker

📌

Hi Product Hunt, I built CGA because AI coding agents often fail before generation: they do not get the right code evidence. CGA indexes a repo into a local graph of files, symbols, calls, imports, dependencies, and lightweight data flow, then exposes MCP-compatible tools so agents can retrieve focused context before editing. The first benchmark covers 102 real-code cases: 90.44% average prompt-token reduction and 13.34% average HPS reduction, with the nuance documented publicly. I would love feedback on MCP retrieval, the graph viewer, install flow, and benchmark design.

Report

3mo ago

Forum Threads

p/context-graph-agent

•

3mo ago

What should CGA measure next for AI coding context?

CGA (Context Graph Agent) is launching today, and I would love technical feedback from people building or using AI coding agents.

The core idea is simple: instead of giving an agent broad file dumps or noisy search results, CGA indexes a repository into a local graph of files, symbols, calls, imports, dependencies, and lightweight data flow. Agents can then query focused evidence through MCP-compatible tools before generating code.

View all